fudged71

7 ms·

61.
by fudged71 10mo ago
I've watched maybe 10 DIY projects of this nature with 3D printing and this is by far the best image quality for the small size of the body. Incredible work
62.
by fudged71 10mo ago
They’re way past being relevant and that is the problem they are solving… getting mindshare again.
63.
by fudged71 10mo ago
This tells me that we need to build 1000 more linters of all kinds
64.
by fudged71 10mo ago
I mean it in a very particular way. The vendor selling the 3D-printed part at an airshow probably didn't think: "I'll deceive pilots." They likely thought: "I can 3D-print this part to spec, it looks right, it fits,
65.
by fudged71 10mo ago
That's insane. I started Ask Me Anything on reddit, does being a moderator in that capacity mean I limited free speech of Americans?
66.
by fudged71 11mo ago
This is the mechanical equivalent of vibe coding. 3D printing itself isn't exactly to blame but the negligence of the company that created and sold this part and omitted it's use from an inspection. Just because a part has the sha
67.
by fudged71 11mo ago
Client CEO was one-shot by a vibecoded startup. Basically the YCombinator dream scenario (for the startup not you). A lot of lessons from this for incumbents. Disrupt yourself before it happens to you.
68.
by fudged71 11mo ago
Is there a centralized tracker for all these off-label benefits of GLP-1? Besides meta-reviews and one-off Deep Research reports
69.
by fudged71 11mo ago
I’m now using it on iOS, switching from Arc. It’s working surprisingly well, good feel and customization. Might keep it.
70.
by fudged71 11mo ago
What’s the license? Might be cool to port to the web with WASM (looks like it would be pretty easy actually)
71.
by fudged71 11mo ago
If I understand correctly, Claude Code Agent SDK can’t edit itself (to incrementally improve), can’t edit or create its own agents and skills. I’ve found that by running Claude Code within Manus sandbox I can inspect the reasoning traces an
72.
by fudged71 11mo ago
I'm currently building on the Auth0 SaaStarter because it seemed to be the only option in the market for something with all the core features enterprises are looking for. Is there an alternative that doesn't require building from
73.
by fudged71 11mo ago
Anyone can be targeted, anyone can be pulled over, anyone can be detained, anyone can be labelled a terrorist, anyone can be deported to a black hole in el salvador. These are dark times.
74.
by fudged71 11mo ago
I’ve been really excited for you infographic generation. Previous models from Google and openAI had very low detail/resolution for these things. I’ve found in general that the first generation may not be accurate but a few rolls of the
75.
Agentic AI's OODA Loop Problem(schneier.com)
1 points by fudged71 11mo ago|0 comments
76.
by fudged71 11mo ago
Extraordinary claims need extraordinary evidence
77.
by fudged71 11mo ago
I think prolog is the right format to codify expertise in Claude Skills. I just haven’t tested it yet.
78.
by fudged71 11mo ago
Homebrew also started preventing you from installing any packages system-wide with pip
79.
by fudged71 11mo ago
I almost exclusively use Deep Research as inputs to LLMs for deeper domain knowledge including frontier scientific theories etc.
80.
by fudged71 1y ago
Sorry, I'm well aware of skills, but wanted to know how specifically they would be used in this Excel extension
81.
by fudged71 1y ago
Interesting their X post mentions "pre-built Agent Skills" but it's not on the webpage. I wonder if they will give you the ability to edit/add/delete Skills, that would be phenomenal. Edit: found it on their other b
82.
by fudged71 1y ago
Huh? Integrating external tool calls for routine, deterministic operations in large language models is not a fallback or workaround but a deliberate, intelligent architectural choice that mirrors human adaptive expertise by combining effici
83.
by fudged71 1y ago
The bar is much lower https://www.nytimes.com/2025/05/27/us/politics/trump-pardon-...
84.
by fudged71 1y ago
The combination of projects, skills, and memory should be really powerful. Just wish they raised the token limits so it’s actually usable.
85.
Gephi Lite v1.0(gephi.wordpress.com)
16 points by fudged71 1y ago|0 comments
86.
by fudged71 1y ago
Maybe it’s a combination of two things (a perfect storm): 1) APFS's snapshot feature, driven by Time Machine, adds a heavy tax on the massive number of file writes and deletes performed by pnpm and git. 2) macOS's less efficient h
87.
by fudged71 1y ago
Most interesting aspects (Notation+DSL) here: Physical Reality as Type System - Knitting notation treats physical constraints (yarn direction, needle capacity) as compile-time invariants, not runtime checks - The notation literally cannot e
88.
by fudged71 1y ago
It really does feel like there is a long term strategy with miniaturization and functional integration. The power of a MacBook Pro in the bump of a phone, the rest is just battery, screen, antennas and heat dissipation. What other form fact
89.
by fudged71 1y ago
Taking a peek at the research: “While flat-optics cameras have transitioned from theoretical concepts to high-fidelity laboratory prototypes, significant interdisciplinary R&D—spanning nanofabrication, materials science, computational i
90.
by fudged71 1y ago
We should chat. We rapidly capture the existing operating models in universities for analysis and optimization of administrative workflows.

More ›